1. Field of the Invention
The present invention relates to an image processing apparatus for reproducing an image.
2. Related Background Art
The present applicant proposed a technique for forming a halftone image at a laser beam printer or the like, wherein an input digital image signal is converted into an angles image signal, and the analog image signal is compared with a periodic analog pattern signal such an a triangle wave signal to produce a PWM image signal.
An example of this technique is shown in FIG. 3. Eight input digital video signal with VD0 to VD7 are latched by a latch circuit 1 in response to a video clock signal CLK/2 obtained by dividing a master clock signal CLK into 1/2 and is synchronized. The latched video signal is converted into an analog video signal VA by a D/A converter 2. The analog video signal VA is input to the noninverting input terminal comparator (CMP) 4.
The master clock signal CLK is frequency-divided by a JK flip-flop (FF) 8 into pattern clock signals PCLK each having a 50% duty ratio. The pattern clock signal PCLK is input to an integrator through a buffer 9. The integrator comprises a variable resistor 10' and a capacitor 11. The pattern clock signal PCLK is integrated into a triangle wave signal (analog pattern signal) SAW having the same period as that of the pattern clock signal PCLK. The bias component of the triangle wave signal SAW is controlled by a capacitor 12 and a variable resistor 13. The triangle wave signal SAW then passes through a protective resistor 14 and a buffer amplifier 15. The triangle wave signal SAW from the buffer amplifier 15 is input to the inverting input terminal of the comparator 4. The comparator 4 compares the level of the analog video signal VA with the level of the triangle wave signal SAW. The analog video signal VA is PWM-modulated according to its density.
In order to obtain good gradation, the amplitude levels of the analog video signal VA and the triangle wave signal SAW preferably satisfy the relationship shown in FIG. 4. More specifically, the maximum level VAmax (e.g., the black level) of the analog video signal VA coincides with the peak level of a triangle wave signal SAW1, and at the same time, the minimum level VAmin (e.g., the white level) of the analog video signal VA coincides with the bottom level of the triangle wave signal SAW1. Therefore, the maximum resolution and linearity in the full scale can be obtained.
However, in practice, even the white level requires a small number of laser beam pulses due to problems associated with the response time of a semiconductor laser element and the sensitivity of a photosensitive drum. Similarly, the black level may not often require a continuous laser beam. In order to obtain a high gradation or tonality output level, a first predetermined pulse width is required even for the white level, and a second predetermined pulse width is required even for the black level. This is mostly due to the response time of the semiconductor laser element, and details thereof are described in U.S. application Ser. No. 51,154, filed May 18, 1987, now U.S. Pat. No. 4,800,442, which is a continuation of U.S. application Ser. No. 931,941, filed Nov. 19, 1986, now abandoned, which is a continuation of U.S. application Ser. No. 765,938, filed Aug. 18, 1985, now abandoned
FIG. 5 is a timing chart for explaining the control technique for the circuit of FIG. 3. Assume that the bias of the triangle wave signal is set to be SAW2 by the variable resistor 13 so as to obtain a small number of laser beam pulses even if D/A input data OO.sub.H (H denotes the hexadecimal notation) for the white level is input. In this case, the white level pulse width is given as w. However, reproducibility of a gray scale level near the black level is degraded in this condition. D/A input data FF.sub.H is input to cause the variable resistor 10' to control the amplitude of the triangle wave signal. The resultant triangle wave signal is SAW3. However, the peak level of the triangle wave signal is also changed. At the same time, the pulse width for the white level is also spread in a direction of arrow a. Therefore, a target value cannot be easily obtained, and a long trial-and-error period is undesirably required.